We propose to develop statistical and computational methods for interpreting data on gene regulation in a variety of cell types from the ENCODE Project. Our work will focus on (i) developing methods for integrating diverse data types in many tissues to infer chromatin architecture; (ii) inferring the sequence determinants and regulatory consequences of differences in chromatin across cell types; and (iii) empirical tests of our predictions using both natural genetic variation and experimental approaches.